JP2018005392A - Formulation support system and formulation support method for operation maintenance knowledge information - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently extract knowledge information on operation maintenance from an existing system.SOLUTION: A formulation support system 1 supports formulation of knowledge information related to operation maintenance by using information extracted from an existing system 4. The formulation support system comprises: an information acquisition part F10 for acquiring predetermined information related to the existing system 4; a key element extraction part F12 for analyzing the acquired predetermined information to extract key element information as a key element related to logical determination of operation maintenance; and an operation maintenance knowledge information extraction part F14 for performing simulation analysis of the extracted key element information to extract and store the operation maintenance knowledge information as knowledge information related to operation maintenance.SELECTED DRAWING: Figure 1

Description

The present invention relates to an operation maintenance knowledge information formulation support system and a formulation support method.

  Conventionally, for example, in various plants such as power plants, steel manufacturing plants, and chemical plants, a specific management company (operator) has been in charge of operation and maintenance for many years, improving the quality and improvement of delivered equipment, etc., and improving operation and maintenance technology We are striving to Management companies that have operated the plant for many years have accumulated unique knowledge and operational methods based on experience.

  By the way, in recent years, there is a tendency to aim at reduction of plant construction cost and operation cost by appropriately combining products purchased from a plurality of vendors. Vendors and products are selected from time to time, without limiting vendors or products. In the past, products provided by vendors with high reliability and a proven track record (especially, custom-made products and dedicated products for specific plants) have been used in many plants. However, in recent years, customers have demanded that an inexpensive plant be constructed by combining inexpensive general-purpose products purchased from various vendors.

  If the new plant is made up of the same products as the existing plant and is operated by the same management company as the operation and maintenance management company of the existing plant, the know-how of operation and maintenance technology cultivated in the existing plant will be used for the new plant. It can be applied almost as it is.

  However, when creating a new inexpensive plant by combining other companies' products and general-purpose products that have not been used successfully, or when a newly entered management company operates a new plant, the know-how of operation and maintenance technology possessed by the existing plant is transferred to the new plant. It is difficult to apply.

  This is because much of the know-how of operation and maintenance technology is not explicit knowledge, but is tacit knowledge. If the know-how of operation and maintenance technology in an existing plant cannot be applied to a new plant, it becomes difficult to stably operate and maintain the new plant. However, it is difficult to extract and formalize know-how of operation and maintenance technology accumulated in existing plants, and it is difficult to predict costs.

  Here, Patent Document 1 as a conventional technique detects an abnormality using an operation state such as an operation time of a plant facility and output signals of a plurality of sensors provided in the plant facility. In Patent Literature 1, a detected abnormality and a countermeasure are linked from past countermeasure examples such as work history and replacement part information to improve diagnosis accuracy.

JP 2013-41448 A

  According to the prior art of Patent Document 1, maintenance work based on the experience and intuition of workers can be automated, and the know-how of operation and maintenance technology that has been tacitized can be formalized. However, Patent Document 1 is a technique that presupposes that the monitoring target and the operation target are clear, and gives a predetermined command to the operation target when the monitoring target satisfies a predetermined condition. Therefore, it cannot be applied to plants with different monitoring targets and operation targets. What should be monitored in the first place? This is because it is unclear what and how to operate.

  Solving this problem becomes more difficult when multiple partners are involved and a system is constructed by combining hardware and software provided by many manufacturers.

  The present invention has been made in view of the above-described problems, and an object thereof is to provide an operation maintenance knowledge information formulation support system that can efficiently extract knowledge information related to operation maintenance from an existing system. .

  In order to solve the above problems, the operation maintenance knowledge information development support system according to the present invention is an operation maintenance knowledge information development support system that supports the development of knowledge information related to operation maintenance using information extracted from an existing system. An information acquisition unit that acquires predetermined information about the existing system, a key element extraction unit that extracts key element information that is a key element related to logical judgment of operation maintenance by analyzing the acquired predetermined information, An operation maintenance knowledge information extraction unit that extracts and stores operation maintenance knowledge information that is knowledge information related to operation and maintenance by performing simulation analysis on the extracted key element information.

  According to the present invention, operation maintenance knowledge information can be extracted by extracting key element information related to logical judgment of operation maintenance from an existing system and performing simulation analysis.

It is explanatory drawing which shows the whole outline | summary of this embodiment. It is a block diagram of a plant operation knowledge formulation support device. It is a block diagram of a client terminal and an operation and maintenance terminal. It is an example of system log data (input data). It is an example of system log data (output data). It is an example of work log data. It is an example of external data. It is an example of key element data. It is an example of knowledge data. It is a flowchart of a knowledge formulation support process. It is an example of log data according to the second embodiment. It is an example of key element data. It is a flowchart which shows the process which concerns on 3rd Example and searches for knowledge establishment conditions. The flowchart of the process which concerns on 4th Example and collects screen operation log. Explanatory drawing in the case of applying the knowledge extracted from the existing plant to a new plant according to the fifth embodiment. Explanatory drawing which shows a mode that it concerns on 6th Example and classifies the knowledge extracted from the existing plant for every customer policy.

  The operation maintenance knowledge information development support system according to the present embodiment has a function for analyzing monitoring records and control records related to the operation and maintenance of a system such as a plant, and an analysis result for each logical judgment and action for control. It has a function to classify, a function to change the frequency related to monitoring and control from the classification result, and a function to simulate plant operation and maintenance based on the same conditions as the analysis result and the changed frequency. Determine the frequency conditions related to monitoring and control necessary for judgment and action.

  In the present embodiment, monitoring data and command data are analyzed using big data analysis technology from three types of operation records of monitoring data (sensor data), logic judgment, and command data related to plant operation and maintenance. Thereby, in the present embodiment, it is possible to extract mechanical feature amounts associated with logical judgment, specifically, specific monitoring data and command data. In the present embodiment, these monitoring data, command data, and logical judgment (detection of an abnormality sign, etc.) are stored as knowledge of operation and maintenance. By applying knowledge extracted from an existing plant to a new plant, the convenience and reliability of operation and maintenance of the new plant can be increased at low cost.

  As a technique for extracting knowledge from key element data, for example, Adaptive Resonance Theory (ART) can be used.

  In this embodiment, by changing the frequency of the feature quantity extracted for each logic judgment, whether or not the logic judgment is possible is confirmed by a simulator, and the feature quantity specifications are determined. For example, the simulation analysis reveals that an abnormal sign can be detected when the number of times a signal from a sensor indicates a value within a predetermined range reaches a predetermined number. In the present embodiment, it is possible to determine the specification of the feature quantity for invoking the logical judgment such as the value and the number of times of the sensor signal.

  According to the present embodiment, it is possible to automatically extract operation and maintenance knowledge (know-how) buried as tacit knowledge in an existing plant and make it formal knowledge. And according to this embodiment, the operation and maintenance knowledge made into formal knowledge can be packaged and applied to the operation and maintenance of a new plant.

  FIG. 1 shows an outline of the present embodiment. FIG. 1 shows an outline of the embodiment to the extent necessary for understanding and implementation of the present invention, and the scope of the present invention is not limited to the configuration shown in FIG.

  The operation / maintenance knowledge information formulation support system 1 includes a computer device and is connected to the client terminal 2 and the operation / maintenance system 3. The formulation support system 1 realizes each function required as a formulation support system by the microprocessor executing a predetermined computer program. FIG. 1 shows main functions F10, F12, F14, etc. among the functions F10 to F16 realized by the formulation support system 1.

  The client terminal 2 is a computer terminal used by a user such as a system administrator, and inputs various data information such as log data to the formulation support system 1 and displays information from the formulation support system 1.

  The operation / maintenance system 3 is a device for performing operation / maintenance of the existing system 4 as a knowledge extraction target, and is a computer terminal. The operation and maintenance system 3 is called an operation and maintenance terminal 3 in the embodiments described later. Monitoring data (also referred to as input data, sensor data, and acquisition data) acquired by the operation and maintenance system 3 from the existing system 4 and command data (also referred to as output data and control data) given from the operation and maintenance system 3 to the existing system 4 are And sent from the operation and maintenance system 3 to the client terminal 2. In some cases, data may be transmitted directly from the operation and maintenance system 3 to the formulation support system 1.

  The existing system 4 is configured as various plants such as a power plant, an iron manufacturing plant, a chemical plant, and a water and sewage plant, for example. The existing system 4 has at least one sensor 41 and device 42.

  Examples of the sensor 41 include a temperature sensor, a pressure sensor, a flow rate sensor, a liquid level sensor, a vibration detection sensor, a rotation sensor, an acoustic sensor, a flame detection sensor, an image sensor, a color sensor, a magnetic sensor, an ammeter, a voltmeter, and a manual. Type switches, photoelectric switches, mechanical limit switches, keyboard switches, touch panels, and the like. Examples of the device 42 include a motor, a turbine, a cylinder, a solenoid, a valve, a pump, a heater, a boiler, and a fan.

  The operation / maintenance system 3 performs operation / maintenance of the existing system 4 on the basis of a predetermined logic judgment (abnormality diagnosis and response). Furthermore, when a user such as an operator detects a sign of abnormality from the behavior of the signal of the predetermined sensor 41, the operation maintenance system 3 gives an instruction to the predetermined device 42. As a result, the detected abnormality sign is quickly dealt with, and the occurrence of the abnormality is prevented in advance. A user's operation on the existing system 4 is stored as a work record in the operation and maintenance system 3 and is also sent to the client terminal 2. Data input from the sensor 41 to the operation / maintenance system 3 and data given to the device 42 from the operation / maintenance system 3 are also sent to the client terminal 2.

  The operation and maintenance knowledge formulation support system 1 extracts operation and maintenance knowledge (hereinafter also referred to as knowledge or knowledge data) as “operation and maintenance knowledge information” by analyzing data related to operation and maintenance of the existing system 4. The extracted operation and maintenance knowledge can be installed in the operation and maintenance system 5 in charge of operation and maintenance of the new system 6.

  The formulation support system 1 holds, for example, processing functions such as a data acquisition unit F10, a key element extraction unit F12, and a knowledge extraction unit F14, and recording data T10-T13, key element data T13, and knowledge data T14. The detailed configuration of the formulation support system 1 will be described later with reference to FIG.

  The data acquisition unit F10 corresponds to an “information acquisition unit”. The data acquisition unit F10 acquires and stores log data T10-T12 as “predetermined information” from the client terminal 2 and / or the operation and maintenance system 3. The data acquisition unit F10 converts data having various data formats into a preset unified data format and stores the data. The data acquisition unit F10 can acquire log data regularly or irregularly. For example, when the data acquisition unit F10 is instructed by a system administrator or the like at a predetermined time, when any maintenance work is performed in the operation maintenance system 3, or when the configuration of the existing system 4 changes, Log data can be acquired and stored at one or more timings.

The key element extraction unit F12 analyzes the log data to extract and store data (key element data) T13 that is a key element related to the logical judgment of operation and maintenance. Here, the logical judgment of the operation and maintenance is the logic for operation and maintenance such as detection of a sign of abnormality and prediction of failure. The logical judgment can include not only pure judgment such as sign detection but also actions such as “setting value change”, “system stop”, and “path switching”.
The key element data T13 as “key element information” is data extracted by performing big data analysis on the log data T10-T12, and is considered to be closely related to the logical judgment of operation and maintenance.

  The knowledge extraction unit F14 obtains knowledge data as “operation maintenance knowledge information” by performing a simulation analysis of the key element data T13. The knowledge extraction unit F14 determines whether there is a relationship between the key element data T13 and the logical determination, and generates and stores knowledge data T14 by associating the related key element data with the logical determination.

  Further, the knowledge extracting unit F14 changes the generation condition of the input data among the key element data by the generation condition changing unit F141. As a result, the knowledge extraction unit F14 confirms what kind of generation condition is used to execute the logical judgment corresponding to the key element data, and adds the generation condition to the knowledge data.

  The knowledge data T14 generated in this manner is obtained by automatically detecting and formalizing the tacit knowledge cultivated in the operation and maintenance work over many years of the existing system 4. At least a part of the operation and maintenance know-how hidden as the experience and intuition of the worker is visualized by the formulation support system 1 and used for the operation and maintenance of the new system 6.

  A first embodiment will be described with reference to FIGS. FIG. 2 is a configuration example of the entire system including the formulation support system 1 according to the present embodiment.

  The overall system includes, for example, a plant operation and maintenance knowledge development support system (hereinafter referred to as knowledge development support system) 1, a plurality of client terminals 2, and a plurality of operation and maintenance terminals 3. It is connected via the communication network CN1.

  An operation and maintenance terminal 3 that performs plant operation and maintenance is connected to a plurality of sensors 41 and devices 42 via a communication network CN2. These sensors 41 and equipment 42 are included in the existing system (existing plant) 4 in FIG. Specifically, the operation and maintenance terminal 3 is connected to a plant maintenance and operation system (not shown), and is a computer terminal that inputs instructions and set values to the plant maintenance and operation system. Hereinafter, for convenience of explanation, the plant maintenance operation system and the operation maintenance terminal 3 may be used without being particularly distinguished.

  The knowledge formulation support system 1 includes, for example, a CPU 11, a memory 12, a communication unit 13, and a storage device 14. Various programs included in the knowledge formulation support system 1 are stored in the storage device 14. The CPU 11 reads and executes these computer programs via the memory 12 as necessary. Thereby, each function F10-F16 is implement | achieved.

  The communication unit 13 is a function for connecting to the communication network CN1 for communication. The knowledge formulation support system 1 communicates with the client terminal 2 and the operation and maintenance terminal 3 via the communication unit 13 and the communication network CN1.

  The storage device 14 is composed of, for example, a hard disk or a flash memory, and stores a computer program for realizing the functions F10 to F16 and data T10 to T14.

  Examples of computer programs stored in the storage device 14 include a data acquisition unit F10, a big data analysis unit F11, a key element extraction unit F12, a simulation analysis unit F13, a knowledge extraction unit F14, a screen control unit F15, and a communication control unit. There is F16.

  The data acquisition unit F10 has a function of acquiring logs of various data formats from the log T20 of the client terminal 2, and converting the acquired data formats into a unified format that can be processed by the knowledge formulation support system 1. . Normally, each client terminal 2 has its own device and system, and the data format is rarely the same. In order to handle a plurality of data having different data formats, the data acquisition unit F10 is required. Note that the function of standardizing the data format can be realized by commercially available software, and thus will not be further described in the present embodiment.

  The big data analysis unit F11 has a function of picking up items that are likely to be related from, for example, three types of logs: system log data T10, work log data T11, and external data T12. Since this function can be realized by using a so-called data mining method and can also be realized by commercially available software, a specific description is omitted.

  The key element extraction unit F12 selects a key element from the related item candidates picked up by the big data analysis unit F11, and stores it in the key element data T13. The pickup result will be described later with reference to FIG. The big data analysis unit F11 normally presents a plurality of related items according to the magnitude of the relationship. For example, the top three related items may be mechanically selected as key elements, or the pickup result may be displayed on the screen, and the user may manually select the key elements from the displayed results.

  The simulation analysis unit F13 is a function for confirming how the plant operation and maintenance system operates from various input data stored in the key element data T13. For the simulation analysis function F13, for example, a training simulator installed in a plant operation and maintenance system can be used. In a training simulator developed for an operator to train with actual equipment, it is possible to output the same result as an actual operation and maintenance operation using data and preconditions used in an actual plant as input sources.

  Based on the result of the simulation analysis unit F13, the knowledge extraction unit F14 allows the combination of the input data and the output data of the key element to operate an advanced automatic logic judgment function such as abnormality detection or abnormality sign detection. Check if you can. When the advanced and automatic logic determination function operates, the knowledge extraction unit F14 stores the input data and output data of the key element data T13 as knowledge data T14.

  The screen control unit F15 creates a screen provided to the user by the knowledge formulation support system 1 and causes the client terminal 2 to display the screen. For example, the screen control unit F15 can display part or all of the contents of the system log data T10, the work log data T11, the external data T12, and the key element data T13 on the screen. Furthermore, the screen control unit F15 can display the analysis result of the big data analysis unit F11 and the analysis result of the simulation analysis unit F13 on the screen. The screen control unit F15 displays the created screen on the client terminal 2 in cooperation with the client execution unit F20 of the client terminal 2. The screen control unit F15 can also detect and save an operation on the display screen.

  The communication control unit F16 has a function of controlling the communication unit 13 to communicate with the client terminal 2 and the operation / maintenance terminal 3. Data transmission / reception between the knowledge formulation support system 1 and each of the terminals 2 and 3 includes not only transmission / reception of log data but also transmission / reception of data used for screen display such as a remote desktop. Further, the communication control unit F16 can preferentially process data transmission / reception with a specific client terminal 2 when the number of client terminals 2 increases, for example.

  A configuration example of the client terminal 2 and the operation and maintenance terminal 3 will be described with reference to FIG. First, the configuration of the client terminal 2 will be described. The client terminal 2 includes, for example, a CPU 21, a memory 22, a communication unit 23, a storage unit 24, and a user interface unit (UI unit in the figure) 25.

  The CPU 21 reads and executes the computer program stored in the storage unit 24 via the memory 22 to realize the functions F20 to F22 required for the client terminal 2. The client terminal 2 is connected to the communication network CN1 via the communication unit 23 and communicates with the knowledge formulation support apparatus 1.

  The storage unit 24 stores a computer program for realizing the client execution unit F20, the screen control unit F21, and the communication control unit F22, log data T20, and knowledge data T21.

  The client execution unit F20 is a function for connecting as a client to the knowledge formulation support system 1 as a server and exchanging information. The client execution unit F20 may be a dedicated program or a web browser.

  The screen control unit F21 is a function for displaying the execution result of each function of the knowledge formulation support apparatus 1 and the log data T20 and the knowledge data T21 held by the client terminal 2 on the user interface unit 25.

  The user interface unit 25 includes an information input unit and an information output unit. The information input unit is a function for a user to input an instruction or the like to the client terminal 2, and includes at least one of a keyboard, a pointing device, a touch panel, a voice instruction device, and the like. The information output unit is a function for outputting information from the client terminal 2 to the user, and includes at least one of a display, a voice synthesizer, a printer, and the like.

  The operation and maintenance terminal 3 includes a CPU 31, a memory 32, a communication unit 33, a storage unit 34, and a user interface unit 35.

  The CPU 31 implements the functions F30 to F32 required for the operation and maintenance terminal 3 by reading and executing the computer program stored in the storage unit 34 via the memory 32. The operation and maintenance terminal 3 is connected to the communication network CN1 via the communication unit 33, and communicates with the knowledge formulation support apparatus 1.

  The storage unit 34 stores a computer program for realizing the operation maintenance unit F30, the screen control unit F31, and the communication control unit F32, log data T30, and knowledge data T31.

  The operation and maintenance terminal 3 is connected to the communication network CN1 via the communication unit 33, and communicates with the knowledge formulation support apparatus 1. Further, the operation and maintenance terminal 3 can communicate with the plurality of sensors 41 and the plurality of devices 42 via the communication network CN2. The operation / maintenance terminal 3 may be connected to the communication network CN2 via the communication unit 33, or may be connected to the communication network CN2 via another communication unit (not shown).

  The operation maintenance unit F30 is a function for performing operation maintenance of the plant 4 based on the knowledge data T31. The operation maintenance unit F30 performs advanced and automatic logic determination such as abnormality detection and abnormality prediction based on the monitoring data acquired from the sensor 41 and the data stored in the knowledge data T31. The operation maintenance unit F30 selects the command data in the knowledge data T31 according to the result of the advanced logic judgment, and instructs the device 42 by the command data. Thus, the operation and maintenance terminal 3 operates and maintains the plant 4 in real time based on the knowledge data T31 obtained from the results of operation and maintenance and the state of the plant.

  The screen control unit F31 has a function of displaying the execution result of the operation maintenance unit F30, the log data T30 related to the plant operation maintenance of the operation maintenance terminal 3, and the acquired knowledge data T31 on the user interface unit 35.

  FIG. 4 shows an example of system log data T <b> 10 </ b> A related to input data among log data collected from the plant 4. The system log data of the input data is included in the system log data T10 shown in FIG.

  The input data is data for acquiring the state of the plant 4 and can also be called, for example, sensor data, monitoring data, or acquisition data. Since the data detected by the sensor 41 is input to the knowledge formulation support system 1 from the plant 4 side, it is referred to as input data (monitoring data, sensor data) in this embodiment.

  The data acquisition unit F10 generates system log data T10A of input data by converting input data of various data formats into data of a unified format.

  The system log data (input data) T10A includes, for example, a date and time C100A representing date and time, an acquired value C101A acquired by the sensor 41, a device name C102A representing a device in which the sensor 41 is installed, and a sensor name representing a sensor name. C103A.

  The acquired value C101A has a measured value C101A1 and a correction value C101A2. The correction value C101A2 is a value when the measurement value C101A1 is converted into a unified format. If the measured value C101A1 is the value that is output from the sensor 41, it can be said that the correction value C101A2 is a value in which the unit and accuracy are unified in order for the knowledge formulation support system 1 to calculate all data in a unified manner.

  FIG. 5 shows an example of system log data T10B related to output data among the system log data T10. The output data is data output to the equipment 42 included in the plant 4 and can also be called command data or control data. The system log data T10B of the output data can also be called actuator command data. The actuator is a device 42 and is an active mechanical element included in the plant 4.

  The system log data (output data) T10B includes, for example, a date and time C100B representing the date and time, a specified value C101B designated for the actuator, a device name C102B representing the device where the actuator is installed, and an actuator name C103B representing the actuator name. including.

  The command value C101B has a designated value C101B1 and a correction value C101B2. The correction value C101B2 is a value when the specified value C101B1 is converted in format. Assuming that the designated value C101B1 is a value as it is, the correction value C101B2 is a value in which units and styles are unified so that the knowledge formulation support system 1 can uniformly calculate all data. Under the unified format, for example, “turn on” is indicated as “1”, and “turn off” is indicated as “0”. The actual command value varies depending on the specifications of the device 42. The actual command value according to the specification of the device 42 is stored in the designated value C101B1.

  FIG. 6 shows an example of work log data T11. The work log data T <b> 11 is work record data related to the operation and maintenance of the plant 4. The data acquisition unit F10 creates work log data T11 by converting various data formats into a unified format.

  The work log data T11 is processed with respect to, for example, the date and time C110 representing the date and time, the work ID C111 which is an identifier indicating what processing work has been performed, the work name C112 indicating the work content, and which device 42 is processed. And the device name C113 indicating whether or not

  FIG. 7 shows an example of the external data T12. The external data T12 is data acquired from an external system such as a weather forecast system outside the knowledge formulation support system 1. The data acquisition unit F10 creates external data T12 by converting various data formats into a unified format.

  The external data T12 includes, for example, a date and time C120 that indicates the date and time, an external data ID C121 that is an identifier that indicates the external data type, a data name C122 that indicates the content, and a value C123 that is the value.

  The external data includes, for example, weather data, operation history data indicating a plant operation date, a maintenance inspection date, and the like, a predicted value output from the production prediction system, and the like.

  FIG. 8 shows an example of the key element data T13. The key element data T13 is created by the key element extraction unit F12.

  The key element data T13 includes, for example, an action ID C130 that represents an identifier of the key element, acquired data C131 that is a source of logic judgment related to operation and maintenance, and command data that indicates the contents commanded to the device 42 as a result of the logic judgment. C132.

  The acquired data C131 includes a device name C1311 indicating from which device 42 the data is acquired and a sensor name C1312 indicating from which sensor 41 the data is acquired. In FIG. 8, only one type of acquired data C131 is associated with one action ID C130, but a plurality of acquired data C131 can be associated with one action ID C130 and stored.

  The command data C132 includes a device name C1321 indicating to which device 42 the command has been issued, and a control target C1322 indicating to which element of the device 42 the control command has been issued. The control target C1322 can also be referred to as a control target element name or an actuator name. In FIG. 8, only one type of command data C132 is associated with one action ID C130, but a plurality of command data C132 may be associated with one action ID C130.

  In FIG. 8, the record with the action ID C130 of “001” is an example in which the big data analysis unit F11 determines that there is a relationship by the mining process. In this case, “none” is stored in both the device name C1311 and the sensor name C1312.

  In FIG. 8, the record with the action ID C130 of “009” is another example that the big data analysis unit F11 determines to be related by the mining process. In this example, the command data C132 is “external cooperation” while the sensor name C1312 of the acquired data C131 is “vibration detection sensor”. In other words, in this example, the device 42 is not instructed directly but is instructed to the external system. For example, when there is a sign of abnormality in the vibration of the turbine, instead of lowering the rotational speed of the turbine, for example, an external maintenance management system (not shown) is requested to replace the equipment 42, Request a schedule adjustment.

  In actual operation, since the operation of the device 42 is confirmed before and after the command, the acquired data C131 can hardly be “None”, but here it is omitted for the sake of simplicity. ing.

  FIG. 9 shows an example of knowledge data T14. The knowledge data T14 is generated by the knowledge extraction unit F14.

  The knowledge data T14 includes, for example, a knowledge ID C140, a logical determination C141, a related key element ID C142, acquired data C143, and command data C144.

  The knowledge ID C140 is an identifier that identifies knowledge related to operation and maintenance of the plant 4. The logic judgment C141 is a judgment that is the content of the knowledge. Examples of logic judgment include detection of abnormality, detection of abnormality sign, and the like. The related key element ID C142 is an identifier that specifies a key element related to the knowledge specified by the knowledge ID C140.

  The acquired data C143 specifies the type of input data (sensor data) that is a premise of knowledge. The acquired data C143 includes a device name C1431 that identifies the device in which the sensor is installed, and a sensor name C1432 that identifies the sensor.

  The command data C144 specifies a destination of command data (output data, control data) as a result of applying knowledge. The command data C144 includes a device name C1441 that specifies a device that gives a command, and a control target C1442 that specifies a target to be controlled by giving the command. The control target C1442 is an element name to be controlled as a result of knowledge among one or a plurality of control target elements included in the device.

  FIG. 10 is a flowchart showing a process for supporting the formulation of knowledge for operation and maintenance. This process is executed by the knowledge formulation support system 1. The user can cause the knowledge formulation support system 1 to execute this process at an appropriate timing.

  The knowledge formulation support system 1 roughly performs the following two types of processing in order. The first process is a process of extracting data as key elements (S10 to S15). The second process is a process for extracting and determining knowledge data (S16 to S20).

  Key element extraction processing (S10 to S15) acquires and analyzes various data for extracting key element data, and stores the analysis results. In the knowledge confirmation process (S16 to S20), the extracted key element data is used to confirm whether or not the advanced logic judgment function is automatically operated, and the confirmed key element data and the like are stored as confirmed knowledge data. Whether or not the key element data automatically executes advanced logical judgment can be confirmed by simulation processing.

  First, the data acquisition unit F10 acquires data necessary for generating knowledge data from the log data T20 stored in each client terminal 2 (S10), and standardizes the format of the acquired data (S11). That is, the data acquisition unit F10 classifies the data collected from each client terminal 2 into system log data, work records, and external data, unifies the format of the classified data, and obtains system log data T10, work log data. Store to T11, external data T12 (S11).

  The big data analysis unit F11 reads data from the system log data T10, the work log data T11, and the external data T12, and analyzes the read data by performing data mining processing (S12). The analysis result is displayed on the user interface unit 25 of the client terminal 2 (S13).

  For example, the key element extraction unit F12 executes step S13 for displaying the analysis result. The key element extraction unit F12 displays the analysis result output from the big data analysis unit F11 on the user interface unit 25 of the client terminal 2 via the screen control unit F15. The user can confirm the big data analysis result. This step S13 may be omitted.

  The key element extraction unit F12 extracts candidate data that can be key element data based on the data extracted in the big data analysis process (S12), and stores the candidate data in the key element data T13 (S14).

  When selecting a candidate for key element data from the results of the big data analysis process, a user's judgment can be added. That is, the user may select one or a plurality of key element data candidates from the screen displayed in step S13. Alternatively, a predetermined number of upper data among the results of the big data analysis process may be automatically selected as candidate data for the key element data.

  The key element extraction unit F12 determines whether all data that can be key element data candidates have been extracted from the result of the big data analysis processing (S15). When all key element data candidates have been extracted (S15: YES), the key element extraction unit F12 proceeds to the next step S16, otherwise (S15: NO), the process returns to step S13, and other analysis results are obtained. judge. The above is the key element extraction process. Following this, a knowledge confirmation process is executed.

  In the knowledge determination process, a step of creating input data for simulation (S16), a step of analyzing simulation (S17), and a step of determining whether or not advanced automatic logic judgment such as abnormality detection is executed (S18). And a step of storing key element data (a set of input data and output data) and logic determination in association with each other as knowledge data (S19), and a step of determining whether all key element data have been analyzed (S20).

  In the step of creating input data for simulation (S16), among the data stored in the key element data T13, input data (sensor data, monitoring data, acquired data) is processed as input data to the simulation analysis process.

  In the simulation analysis step (S17), when the simulation analysis unit F13 incorporates the input data created in step S16 and advanced automatic logic judgment such as abnormality detection into the plant operation and maintenance system as knowledge data, Analyzes how the plant operation and maintenance system operates and outputs the result. In step S16, a training simulator provided in advance for plant operation and maintenance may be used, or a simulator using a simple model may be used. Alternatively, the user may consider the operation of the operation and maintenance system on the desk.

  In the step of determining whether or not advanced automatic logic judgment such as abnormality detection is executed (S18), the knowledge extraction unit F14 executes advanced automatic logic judgment such as abnormality detection from the result in the simulation analysis unit F13. Check if it was done.

  When the knowledge extraction unit F14 can execute the advanced automatic logic judgment (S18: YES), the key element data and the logic judgment used for the simulation analysis process are stored as a part of the knowledge data T14 (S19). ).

  On the other hand, when the knowledge extraction unit F14 cannot execute advanced automatic logic determination (S18: NO), it is determined whether simulation analysis has been performed for all key element data (S20). If there is key element data that has not been analyzed yet (S20: NO), the process returns to step S16. When the simulation analysis is performed for all the extracted key element data (S20: YES), this process is terminated.

  According to this embodiment configured as described above, it is possible to automatically extract the knowledge data that has been tacitized in the existing plant 4.

  According to the present embodiment, at least part of the operation and maintenance know-how hidden as the experience and intuition of the worker can be visualized by the knowledge formulation support system 1 and can be utilized for the operation and maintenance of the new plant 6. It becomes.

  Therefore, if knowledge data that has been formalized using this embodiment is installed as software in a new plant operation and maintenance system, even a new business operator who does not have advanced plant operation and maintenance know-how High operational maintenance can be performed. Moreover, by applying the knowledge data formalized by this embodiment to a plant that combines products of other companies whose characteristics and results are unknown, stable and high-quality operation and maintenance can be provided.

  In this embodiment, as an example of knowledge, an advanced automatic determination function such as abnormality detection has been described. However, knowledge is not limited to detection of abnormality or detection of abnormal signs. All functions related to operation and maintenance, that is, a series of specific processes related to the judgment of a machine and a person related to a specific sensor 41 and a specific device 42 are knowledge.

  A second embodiment will be described with reference to FIGS. Each of the following embodiments, including the present embodiment, corresponds to a modification of the first embodiment, and therefore, differences from the first embodiment will be mainly described. In this embodiment, a case where the knowledge is expanded will be described. In the present embodiment, the log data of the plant 4 includes logical judgments and actions.

  FIG. 11 shows an example of a logical judgment / action table T10C included in the system log data T10 of this embodiment.

  The logical judgment / action table T10C includes, for example, a date / time C100C representing the date / time, an execution name C101C indicating what kind of processing has been performed, a factor name C102C indicating an event that caused the processing, and a subject C103C that is the execution subject of the processing The target device C104C indicating to which device 42 the subject C103C has performed processing is included.

  There are two types of the main body C103C: a case of “machine” and a case of “person”. The case where the main body C103C is “machine” is a case where processing is automatically performed based on a predetermined condition, and human judgment is not included in the processing. For example, when a certain set value is exceeded, the process is such that the switch is automatically turned off.

  On the other hand, the case where the subject C103C is “person” is a case where processing is performed through judgment by a person. For example, it is processing that must be processed by human judgment or processing for sudden failure or abnormality.

  The big data analysis unit F11 of this embodiment is an item that is likely to be related to an action through mechanical logic judgment or human judgment from three types of log data of system log data T10, work log data T11, and external data T12. To pick up. Since this function can be realized by using a so-called data mining method, further explanation is omitted.

  FIG. 11 shows an example of the key element data T13_1 of the present embodiment. The key element data T13_1 of the present embodiment further includes the contents of the logic judgment / action table T10C described in FIG. 10 with respect to the key element data T13 described in FIG.

  That is, the key element data T13_1 includes an action ID C130 that indicates identification of the key element, an action C133 that indicates what kind of logical judgment and action has been performed, and acquisition data C131 that indicates the content acquired to reach the action. And command data C132 indicating the contents commanded to cause the action.

  The acquired data C131 includes a device name C1311 that indicates from which device 42 the data has been acquired, and a sensor name C1312 that indicates from which sensor 41 the data has been acquired. In this embodiment, only one type of acquired data C131 is associated with one action C133. However, the present invention is not limited to this, and a plurality of acquired data C131 may be associated with one action C133.

  The command data C132 includes a device name C1321 indicating to which device 42 the command has been issued, and a control target C1322 indicating to which control target the device 42 has issued the control command. In this embodiment, only one type of command data C132 is associated with one action C133. However, the present invention is not limited to this, and a plurality of command data C132 may be associated with one action C133.

  Attention is focused on the record having the action ID C130 of FIG. This shows a case where the action C133 is “start operation” and the operation of the device 42 is started. Here, in order to explain that sensor data is not necessary at the start of operation, no value is set in the acquired data C131. In the device name C1311 and the sensor name C1312, “none” is stored.

  On the other hand, in order to start the operation of the device 42, the command data C132 is necessary. For this reason, values for specifying the target of the control command are set in the device name C1321 and the control target name C1322 of the command data C132.

  In actual operation, since the operation of the device 42 is confirmed both before and after the command, there is almost no case where “none” is set in the acquired data C131, but here the description is simplified. It is omitted to do.

  The key element data T13_1 includes the following five types according to the value acquisition destination and the value setting method. In the following description, the case where no value is set means that “none” is set in FIG. 12.

  In the first case, no value is set in the acquisition data C131 according to the action C133, and values are set in the device name C1321 and the control target name C1322 of the command data C132, respectively.

  The second case is a case where values are set in the device name C1311 and the sensor name C1312 of the acquired data C131, and no value is set in the command data C132.

  The third is a case where values are set in the acquired data C131 and the command data C132, respectively.

  The fourth case is a case where “external cooperation” is set in the device name C1311 of the acquired data C131. In this case, by acquiring data from the external system, it cooperates with the external system.

  The fifth is a case where “external cooperation” is set in the device name C1321 of the command data C132. In this case, by issuing a command to the external system, it cooperates with the external system.

  If the knowledge formulation support process (FIG. 10) described in the first embodiment is executed using the key element data 13_1 shown in FIG. 12, in addition to the high-level automatic logic judgment such as abnormality detection, a general logic judgment Can support the formulation of knowledge data.

  Configuring this embodiment like this also achieves the same operational effects as described in the first embodiment. Furthermore, according to the present embodiment, a wide range of knowledge including both general logic judgment and advanced automatic logic judgment can be acquired from the existing plant. Therefore, by applying this wide knowledge to the operation and maintenance system of a new plant, even if it is a new business operator who does not have know-how or a new plant is constructed by gathering unfamiliar sensors and equipment, reliable operation is possible. Maintenance can be realized.

  A third embodiment will be described with reference to FIG. In this embodiment, the frequency of acquiring key element data (data acquired from the sensor 41 and data commanded to the device 42), which is necessary for obtaining an advanced and automatic logic judgment such as abnormality detection, is specified. . That is, in this embodiment, in order to realize operation and maintenance knowledge, the performance that the sensor 41 and the device 42 should have is examined.

  FIG. 13 is a flowchart of processing for searching for a condition for establishing knowledge. This process is referred to as a knowledge establishment condition search process. This process is executed by the knowledge extraction unit F14.

  As will be described later, the knowledge establishment condition search process includes detection of a time zone (start time and end time) relating to key element data (S30), frequency change of command data or sensor data (S31), and simulation input data. Creation (S32), simulation analysis (S33), determination of whether automatic judgment function or control based on automatic judgment has been executed (S34), and key element data conditions (upper limit / lower limit, etc.) are saved (S35) ), And a determination as to whether the analysis has been completed for all key elements (S36).

  In step S30, a time zone related to the key element data is set. The time zone is defined by a start time and an end time, for example. In this step S30, the ID C142 of the related key element data is checked from the knowledge data T14, and the ID C130 of the key element data T13 is searched using the ID C142.

  In step S30, the system log data T10A and T10B are searched for the acquired data C131 (sensor data) and the command data C132 related to the knowledge, and the start time and the end time corresponding to the acquired data C131 and the command data C132 are obtained. Decide each. That is, in step S30, for the data (acquisition data and command data) from which the knowledge is extracted, the time zone in which the data is detected is specified.

  In step S31, the acquisition frequency of command data (output data) or acquisition data (input data) is changed. In step S31, the occurrence frequency (detection frequency) of the command data and the acquisition data is calculated in the time period set in step S30, and a frequency that is lower than the latest frequency by a predetermined value is set (acquisition frequency = latest frequency). Acquisition frequency-predetermined value).

  In step S32, input data used for the simulation process is created. In step S32, related data (command data C101B and acquisition data C101A) is acquired from the log data T10A and T10B based on the time zone of the key element data set in step S30 and the acquisition frequency set in step S31. To do. In step S32, the frequency of the data is adjusted to the frequency set in step S31.

  In step S33, the simulation analysis unit F13 analyzes how the plant operation and maintenance system operates from various input data created in the simulation input data creation (S32), and outputs the result. As this simulator, a training simulator for plant operation and maintenance may be used, a simulator utilizing a simple model may be used, or desktop examination may be used.

  In step S34, it is determined whether the automatic determination function or the control based on the automatic determination is executed even when the data acquisition frequency is decreased. In step S34, it is confirmed from the analysis result in the simulation analysis unit F13 whether the automatic determination function or the control based on the automatic determination is executed. If they are executed (S34: YES), the process returns to step S31, and the frequency is further reduced and reanalyzed (S32, S33). If they are not executed (S18: NO), the process proceeds.

  In step S35, the conditions for reproducing the knowledge (the upper limit value and the lower limit value of the acquisition frequency) are associated with the key element data and stored in the knowledge data T14.

  In step S36, it is determined whether the reproduction conditions (conditions for executing control based on the automatic determination function or automatic determination) have been analyzed for all knowledge data. If there is unanalyzed knowledge (S36: NO), the process returns to step S30. If all the knowledge has been analyzed (S36: YES), this process is terminated.

  Configuring this embodiment like this also achieves the same operational effects as the first embodiment. Furthermore, according to the present embodiment, it is possible to determine the frequency of acquiring data from a sensor and the frequency of command to a device (actuator) necessary for knowledge. According to the present embodiment, it is possible to determine the specifications of the sensor 41 and the equipment 42 necessary for making use of the knowledge in the new plant 6, and therefore, operation and maintenance with less waste compared to the first embodiment can be realized. This embodiment can be combined with either the first embodiment or the second embodiment.

  A fourth embodiment will be described with reference to FIG. In the present embodiment, the user's operation history for the screen provided to the client terminal 2 by the knowledge formulation support system 1 is automatically extracted, and the operation history is used as at least part of the work log data T11.

  FIG. 14 is a flowchart of processing for extracting and storing a user operation history for a screen displayed on the client terminal 2. This processing is executed by the knowledge formulation support system 1 in cooperation with the client terminal 2 and the operation / maintenance terminal 3, for example. Hereinafter, an example will be described in which the knowledge formulation support system 1 acquires an operation on the display screen of the operation and maintenance terminal 3 and stores it as a screen operation log.

  The user interface unit 35 of the operation and maintenance terminal 3 displays a screen related to operation and maintenance (S40). The user operates by clicking a button displayed on the screen. The user operation on the screen is detected by the user interface 35 of the operation and maintenance terminal 3 and sent to the knowledge formulation support system 1 via the communication unit 23 (S41).

  When receiving the screen operation data from the operation and maintenance terminal 3 (S50), the screen control unit F15 of the knowledge development support system 1 analyzes the screen operation data (S51). The data acquisition unit F10 stores the analysis result in the screen control unit F15 as the screen operation log in the screen operation log data T15 (S52).

  The screen operation log data T15 manages, for example, an operation ID, an operation date and time, an operation type, and an operation content in association with each other. User IDs may be further associated with each other and managed. By referring to the screen operation log data T15, it is possible to confirm what operation the user has performed on which screen element (button, input field, etc.).

  The knowledge extraction unit F14 uses the screen operation log data T15 as at least part of the work log data T11. Thereby, the knowledge extraction part F14 extracts the key element data which can comprise a knowledge.

  Configuring this embodiment like this also achieves the same operational effects as the first embodiment. Further, in this embodiment, the user's screen operation is automatically obtained and analyzed, and screen operation log data T15 which is a part of the work log data T11 is created. Therefore, it is not necessary for the user to manually record work, and usability is improved. This embodiment can be combined with any of the first to third embodiments.

  A fifth embodiment will be described with reference to FIG. In this embodiment, when the knowledge extracted from an existing plant is applied to the operation and maintenance of a new plant, the viewpoint of the system configuration of the plant and the policy of the customer who operates the plant are considered.

  FIG. 15 shows an overall outline of the present embodiment. A plurality of knowledge is extracted from each of the existing plants P1 to P3. It is assumed that knowledge N10 to N1n is acquired from the plant P1, knowledge N20 to N2n is acquired from the plant P2, and knowledge N30 to N3n is acquired from the plant P3.

  It is not always possible to apply all the knowledge extracted from each of the existing plants P1 to P3 to the new plant P4. The new plant P4 and the existing plants P1 to P3 have different system configurations, and there may be different policies of customers who operate the plants. Therefore, in this embodiment, it is determined whether the knowledge extracted from the existing plant can be applied to the new plant by comparing the configuration of the existing plant with the configuration of the new plant.

  Further, in the present embodiment, it is determined whether or not the knowledge determined to be compatible in the system configuration matches the policy of the customer who operates the new plant. In the case of FIG. 15, only knowledge N1n, N21, N30, and N32 that satisfy both the adaptability in the system configuration and the customer policy is applied to the new plant P4.

  According to this embodiment configured as described above, knowledge applicable to the new plant P4 can be selected from the knowledge extracted from the plurality of existing plants P1 to P3, and can be applied to the new plant P4. This embodiment can be combined with any of the first to fourth embodiments.

  A sixth embodiment will be described with reference to FIG. In the present embodiment, the knowledge extracted from the existing plants P5 to P7 is classified and stored for each prepared policy. In the example illustrated in FIG. 16, it is assumed that knowledge N50 to N5n is acquired from the plant P5, knowledge N60 to N6n is acquired from the plant P6, and knowledge N70 to N7n is acquired from the plant P7. Among these knowledge, knowledge N51, N52, N60, N62, N70, and N72 are classified into the policy of “safety priority”. Other knowledge N53, N5n, N63, N71, N73, and N7n are classified into the policy of “cost priority”.

  According to the present embodiment configured as described above, the knowledge extracted from the existing plant is classified and stored for each policy, so the knowledge according to the wishes of the customer operating the new plant is selected and applied to the new plant. You can also. This embodiment can be combined with any of the first to fourth embodiments.

  In addition, this invention is not limited to embodiment mentioned above. A person skilled in the art can make various additions and changes within the scope of the present invention. In the above-described embodiment, the present invention is not limited to the configuration example illustrated in the accompanying drawings. The configuration and processing method of the embodiment can be changed as appropriate within the scope of achieving the object of the present invention.

  Moreover, each component of the present invention can be arbitrarily selected, and an invention having a selected configuration is also included in the present invention. Further, the configurations described in the claims can be combined with combinations other than those specified in the claims.

  1: operation maintenance knowledge development support system, 2: client terminal, 3, 5: operation maintenance system (operation maintenance terminal), 4: existing system (existing plant), 6: new system (new plant), 41: sensor, 42 : Device, T10: System log data, T11: Work log data, T12: External data, T13: Key element data, T14: Knowledge data, F10: Data acquisition unit, F11: Big data analysis unit, F12: Key element extraction unit F13: Simulation analysis unit, F14: Knowledge extraction unit, F15: Screen control unit, F16: Communication control unit

Claims (12)

An operation and maintenance knowledge information development support system that supports the development of knowledge information related to operation and maintenance using information extracted from an existing system,
An information acquisition unit for acquiring predetermined information about the existing system;
By analyzing the acquired predetermined information, a key element extraction unit that extracts key element information that is a key element related to logical judgment of operation and maintenance;
An operational maintenance knowledge information extraction unit that extracts and stores operational maintenance knowledge information that is knowledge information related to operational maintenance by performing simulation analysis on the extracted key element information;
Support system for formulating operation and maintenance knowledge information. The information acquisition unit acquires, as the predetermined information, input information input to devices constituting the existing system, output information for instructing output to the devices, and work record information related to operation and maintenance,
The key element extraction unit extracts information corresponding to the logical determination from the input information and the output information as the key element information,
The operation and maintenance knowledge information extraction unit confirms whether or not there is a relationship between the key element information and the logic judgment by the simulation analysis, and there is a relationship between the key element information and the logic judgment. If it can be confirmed, the key element information and the logical judgment are extracted as the operation maintenance knowledge information.
The operation support knowledge information formulation support system according to claim 1. The operation / maintenance knowledge information extraction unit determines whether output information included in the key element information appears by performing the simulation analysis while changing a generation condition of input information included in the key element information, and the output The generation condition of the input information necessary for the appearance of information is extracted as a part of the operation and maintenance knowledge information, and stored in association with the key element information and the logical judgment.
The operation support knowledge information formulation support system according to claim 2. The operation maintenance knowledge information extracted by the operation maintenance knowledge information extraction unit is output for application to a new system.
The operation support knowledge information formulation support system according to claim 3. The logical judgment includes a judgment regarding a sign of failure occurrence,
The operation support knowledge information formulation support system according to claim 5. Based on the difference in system configuration between the existing system and the new system, it is determined whether the operation / maintenance knowledge information is applicable to the new system, and only the operation / maintenance knowledge information determined to be applicable is transferred to the new system. Apply,
The operation support knowledge information formulation support system according to claim 5. Of the operation and maintenance knowledge information determined to be applicable to the new system, only the operation and maintenance knowledge information that conforms to the policy associated with the new system is applied to the new system.
The operation support knowledge information formulation support system according to claim 6. Classifying the operation and maintenance knowledge information into a plurality of preset policies;
The operation support knowledge information formulation support system according to claim 5. The predetermined information includes screen operation history information that is a history of operations on a screen for operating and maintaining the existing system.
The operation support knowledge information formulation support system according to any one of claims 1 to 8. An operation maintenance knowledge information development support method that uses a computer to support the creation of knowledge information related to operation and maintenance using information extracted from an existing system,
Obtaining predetermined information about the existing system;
By analyzing the acquired predetermined information, key element information that is a key element related to logical judgment of operation and maintenance is extracted,
By analyzing the extracted key element information by simulation, operation maintenance knowledge information that is knowledge information related to operation and maintenance is extracted and stored.
How to support the creation of operational maintenance knowledge information. The predetermined information includes input information input to devices constituting the existing system, output information for instructing output to the devices, and work record information related to operation and maintenance,
Among the input information and the output information, information corresponding to the logical judgment is extracted as the key element information,
If the simulation analysis confirms whether a relationship exists between the key element information and the logic judgment, and if the relationship exists between the key element information and the logic judgment, Key element information and the logical judgment are extracted and stored as the operation and maintenance knowledge information.
The operation support knowledge information formulation support method according to claim 10. A computer program for causing a computer to function as an operation and maintenance knowledge information development support system that supports the development of knowledge information related to operation and maintenance using information extracted from an existing system,
On the computer,
An information acquisition unit that acquires, as predetermined information, input information that is input to a device that constitutes the existing system, output information that instructs output to the device, and work record information related to operation and maintenance;
A key element extraction unit that extracts key element information that is a key element related to logical judgment of operation and maintenance from the input information and the output information by analyzing the acquired predetermined information;
When the relationship between the key element information and the logical judgment is confirmed by simulation analysis, and when it is confirmed that the key element information and the logical judgment exist, the key An operation / maintenance knowledge information extracting unit that extracts and stores element information and the logical judgment as the operation / maintenance knowledge information, and further performs the simulation analysis while changing the generation condition of input information included in the key element information Thus, it is determined whether the output information included in the key element information appears, the generation condition of the input information necessary for the output information to appear is extracted as a part of the operation maintenance knowledge information, An operation and maintenance knowledge information extraction unit that stores the key element information and the logical judgment in association with each other;
A computer program that realizes each.

Images